Cardionetics Ambulatory ECG Monitors

Evaluation of the C.Net2000+ in General Practice

A year-long planned, open, prospective evaluation was carried out in two centres, in Berkshire and Lancashire, to examine the practicality of undertaking 24-hour ambulatory cardiac arrhythmia testing in general practice using the Cardionetics C.Net2000+ ambulatory ECG monitor, as an alternative to the existing practice of referral to a secondary care hospital.

Trial Protocol

Fifty-three GPs in 16 surgeries participated in the trial over a one-year period and a total of 190 patients were recruited. Inclusion criteria for patients were those exhibiting symptoms indicative of cardiac arrhythmia, including palpitations, light-headedness, dizzy spells, and fainting.

For each of the patients, the GP first considered whether a referral to a hospital cardiology department was required (the initial assessment). All patients then underwent a practice-based 24-hour ambulatory arrhythmia test using the C.Net2000+, and a revised assessment of whether to refer was then made, as a result of the test report from the ECG monitor.

All patients were followed up after a 3-month interval and their condition reviewed to determine whether the decision to refer or not was correct.

Patient Statistics

Patients meeting the inclusion criteria showed symptoms indicative of arrhythmias. The majority (137 patients) had no history of cardiac problems. Table 1 shows the demographic characteristics of the patient group and Table 2 their symptoms and cardiac history.

^†Totals exceed 190, as patients presented with multiple symptoms and complex histories.

Results

The referral decisions based on the initial assessment are shown in Table 3. These results show that prior to an arrhythmia test being undertaken by the GP, 33.7% of patients would not have been referred. According to the revised assessments, a fifth of these (6.3% of the total) were false negatives having undiagnosed arrhythmia, and of these patients, two-thirds (4.2% of the total) needed urgent attention. In addition, 50% of patients would have been referred to a secondary care cardiology department but did not actually require cardiac treatment, so could have been treated locally. Detailed results of patient outcomes are presented in Table 5.

The referral decisions based on the revised assessment after arrhythmia monitoring with the C.Net2000+ are shown in Table 4.

Table 5 presents a summary of the results. The error matrix at the right of the Table identifies true positives, true negatives, false positives and false negatives (TP, TN, FP, FN). The true positives column shows that the decision to refer (either routinely or urgently) after the initial assessment was correct for 31 patients. The 52 true negatives were initially determined to not require referral, and this was confirmed by the revised assessment. The 95 patients in the false positives column would have been referred had it not been for the revised assessment. The false negatives column shows that of the 64 patients who would not have been referred after the initial assessment, according to the revised assessment 12 had not been correctly diagnosed, with 8 needing a routine referral and 4 an urgent referral.

^†TP: True positive, TN: True negative, FP: False positive, FN: False negative

The results of a follow-up with the 143 patients who were not referred are shown in Table 6. A number of patients exhibited symptoms of benign arrhythmias and were reassured. Other patients were treated locally, including changing or adjusting medication to alleviate symptoms, including cardiovascular drug treatments.

Five patients were referred to other (non-cardiac) disciplines; these included neurological and thyroid referrals, and smoking cessation and caffeine control programmes.

The decision not to refer a patient who had previously had a coronary artery bypass graft (CABG) is considered a false negative, although in the 3-month follow-up period no outcome could be determined as the patient had not yet been given an outpatient department appointment.

Publication of Interim Trial Results

Early results from the trial, from the first 73 consecutive patients' data, were published in the British Journal of Cardiology [1]. A number of key messages were highlighted in this article:

• Automated arrhythmia monitoring allows GPs to make informed decisions about referrals.
• Patients with life-threatening arrhythmia are identified more quickly.
• Patients with benign arrhythmia can be reassured and treated locally.
• Use of the C.Net2000+ in primary care can reduce referrals of arrhythmia patients to hospital-based cardiology clinics by about 60%.

Conclusions

This study has shown that automated arrhythmia monitoring in general practice is a means of improving the level of service provided to patients. The number of false positive referrals can be significantly reduced, together with a small increase in urgent referrals. In addition, by eliminating the likelihood of the symptoms being cardiac in origin, other possible causes of the symptoms can be investigated earlier.

Patients without significant arrhythmia can be reassured or treated locally. In doing so, the load on cardiology departments can be reduced by up to 73%, which would reduce waiting lists and assist cardiologists in identifying at-risk patients earlier.

Patients with arrhythmia, who would otherwise have to wait for a cardiology outpatient department appointment, benefit significantly by having their condition diagnosed more quickly.

References

[1] P. Standing, M. Dent, A. Craig, B. Glenville, Changes in referral patterns to cardiac outpatient clinics with ambulatory ECG monitoring in general practice, British Journal of Cardiology, 2001, Vol. 8 No. 6, pages 394–398.